1. Field of the Invention
The present invention generally relates to the field of wireless communication systems. More specifically, the invention relates to downlink, i.e. from the base station to a set of terminal units, mixed voice and data transmission for code division multiple access communication systems.
2. Related Art
In a code division multiple access (“CDMA”) communication system, such as IS-95, or CDMA2000, or WCDMA (wideband CDMA), transmission can be provided for voice communication and data communication simultaneously by transmitting voice and data signals across one or more communication channels. Certain types of signal transmission, for example, voice and certain types of low data rate data transmissions are degraded by delays in transmission. Certain types of data signal transmission, on the other hand, are tolerant of delays in transmission. For example, because the data is tolerant of delay, the data can be grouped into packets and scheduled for transmission. Furthermore, a delayed packet need not be dropped, and transmission errors can be corrected by simply retransmitting a packet at a later time, i.e. rescheduling the packet. Large amounts of packet data information can be transmitted efficiently in short “bursts” of data at high power and high data rate. Conventional voice/data transmission treats voice and data communications similarly by setting up a communication link at a pre-determined data rate, and attempting to transmit voice and data information without exceeding a certain frame error rate. With conventional low data rate voice/data transmission, changes in the data rate generally do not involve significant changes in the overall transmit power; this is because low data rate connections only use a fraction of the total power available at the base station. By way of contrast, transmission of high speed packet data may require frequent extreme changes in data rate which typically involve large changes in power level. Since high data rate transmission uses a significant fraction of the total base station transmit power, the overall base station transmit power level could be significantly affected by the variation in the power used for high data rate transmissions.
In the present application, voice signal transmission and other signal transmission which is degraded by delays in transmission, as well as conventional data transmission where changes in the data rate are infrequent and relatively minor and changes in transmission power levels are small relative to the total base station transmit power, are referred to as “voice”. Data signal transmission, such as high speed packet data, which can be tolerant of delays in transmission and can be scheduled, and typically is transmitted in short “bursts” at high power and high data rate, as well as any signal transmission where changes in the data rate are frequent and extreme and changes in transmission power levels are relatively large, are referred to as “data”.
In order to efficiently accommodate these different types of signal transmission simultaneously, i.e., mixed voice and data transmission, different approaches may be followed. One approach is to specify a different part of the frequency spectrum, i.e. a different “band” of frequencies or frequency band, for each type of signal. Another approach is to multiplex the voice and data signals together through time division. With the time division approach, some of the time available for transmitting the signals is allotted to voice signals and some of the time available for transmitting the signals is allotted to data signals. For example, in a GSM+GPRS system (Global System for Mobile combined with Generalized Packet Radio System) some time slots normally used for regular GSM voice transmission are instead used for packet data transmission. One approach, used as an example in the present application, is code division multiple access (CDMA), which allows multiple signals to be transmitted at the same time on the same frequency band.
In CDMA systems each user's signal is separated from other users' signals by modulating the transmission signal with a distinct spreading code sequence. The modulation of the transmission signal spreads its spectrum so that the bandwidth of the encoded transmission signal is much greater than the original bandwidth of the user's information. For this reason CDMA is also referred to as “spread spectrum” modulation or coding. Each user uniquely encodes its information into a transmission signal using the spreading code sequence. The intended receiver, knowing the spreading code sequence of the user, can decode the transmission signal to recover the information.
By way of background, in CDMA communications, the user's signal is spread to allow many users to simultaneously use the same bandwidth without significantly interfering with one another. One means of spreading is the application of distinct “orthogonal” spreading codes or functions, such as Walsh functions, to each user's signal. “Orthogonality” refers to lack of correlation between the spreading functions. In a given spread spectrum communication system using Walsh functions (also called Walsh code sequences), a pre-defined Walsh function matrix having n rows of n chips each is established in advance to define the different Walsh functions to be used to distinguish different user's signals. As an example, for a given sector (or cell in the WCDMA terminology), each downlink channel is assigned a distinct Walsh function. In other words, communications between a base station and each user are coded by a distinct Walsh code sequence in order to separate each user from the others.
The base station transmits signals to all users in a sector so that the Walsh codes are time synchronized in order to achieve orthogonality between the different signals. Effectiveness of the orthogonal spreading codes is affected by the phenomenon of “multipath”. Simply stated, multipath is interference caused by reception of the same signal over multiple paths, that is, multiple copies of the signal arrive after different path delays. Due to the loss of time synchronization, the orthogonality between different user signals is lost. Interference due to loss of orthogonality through multipath can be averaged by the use of other types of spreading codes such as pseudo-noise (“PN”) sequences, for example. The autocorrelation properties of PN sequences can be used to improve rejection of multipath interference. However, due to the loss of orthogonality through multipath, there is greater interference between the signals of different users, referred to as “intra-cell interference”, including interference of a user's own signal with itself, also referred to as “self-interference”.
In a multi cell system, there can be interference caused by user signals transmitted by the base station in one cell interfering with the user signals transmitted in another cell, also referred to as “inter-cell interference”. The transmit power of the base station transmitters is controlled so as to minimize the amount of power transmitted into neighboring cells in order to limit inter-cell interference. Extreme fluctuations in transmit power can exacerbate the effects of inter-cell interference, as well as intra-cell interference between users including self-interference, described above.
FIG. 1 illustrates an example of the effect of data transmission on power control for multiple voice and data users within the same cell in a CDMA or spread spectrum communication system. FIG. 1 shows graph 100, having power axis 101 plotted against time axis 102. The transmit power for a typical voice user varies in time according to single user voice power curve 104. The aggregate transmission power for all the voice users within the cell is shown as Pv 106 in graph 100. Aggregate voice power Pv 106 varies in time as shown in graph 100. Power is allocated in addition to aggregate voice power Pv 106 for data burst transmissions 108, 109, and 110. The maximum available signal transmission power that can be allocated for the total of aggregate voice and data signal transmissions is maximum power limit Pmax 112, shown in graph 100 as a horizontal solid line and also indicated by “Pmax.” The data and voice aggregate transmission power is shown as Pv+d 114 in graph 100. Data and voice aggregate power Pv+d 114 within the cell varies in time as shown in graph 100. As seen in graph 100, Pv+d 114 remains below maximum power limit Pmax 112.
FIG. 1 shows an example of the effect that data signal transmission can have on power control for a single user in terms of changes to single user voice power curve 104. As a result of data burst transmission 108, interference can be increased, due to the intra-cell effects outlined above, for the single user whose power allocation is represented by single user voice power curve 104. To balance the increased interference, power allocation can be increased for the single user leading to local power peak 105 in single user voice power curve 104. In a conventional voice/data transmission system, changes in power allocation between users tend to balance out, by occurring randomly in time, leaving only a minor effect on aggregate voice power Pv 106. However, the effect of data burst transmission 108 is simultaneous for many users in the cell, so there is a relatively large effect on aggregate voice power Pv 106, shown as increase 116 in aggregate voice power Pv 106.
Continuing with FIG. 1, at the end of data burst transmission 108, interference is reduced for the users within the cell. Thus, the power control system at the base station will decrease the power allocation to the users, leading to decrease 117 in aggregate voice power Pv 106. In a mixed voice and data communication system, the power control system must be able to respond quickly to changes in interference. Thus, decrease 117 in aggregate voice power Pv 106 may be more than needed in view of subsequent data burst transmission 109. In other words, the reaction of the base station's power control system leading to decrease 117 “undershoots” the equilibrium value for stable system performance. As a result, then, of data burst transmission 109, which again causes an increase in interference for the users, the base station's power control system increases the power allocation for the users, leading to increase 118 in aggregate voice power Pv 106. Once again, increase 118 in aggregate voice power Pv 106 may be more than needed. In other words, the reaction of the power control system leading to increase 118 “overshoots” the equilibrium value for stable system performance.
Thus, as shown in FIG. 1, when data signal transmission is mixed with voice signal transmission in a wireless communication system, the different signal characteristics of voice and data transmissions lead to problems with power control for users within the same cell. The signal characteristics of data communications, namely that data transmission typically occurs in bursts, tends to cause disruptions in power control which do not occur with the relatively continuous signal characteristics of voice communications. For example, over-allocation and under-allocation of power to each user and to the aggregate of all users within a cell can disrupt communications and severely degrade the quality of the communication links. In addition, the system becomes subject to large swings in the total power output, as shown by the large variations in the level of data and voice aggregate power Pv+d 114, which indicates the total power output of the system.
FIG. 2A, FIG. 2B, and FIG. 2C illustrate an example of some of the effects of data transmission on power control for users in neighboring cells in a CDMA or spread spectrum communication system. FIG. 2A shows a diagram of cells for exemplary cellular spread spectrum communication system 200 comprising several cells including cell 203, labeled “cell #0” and cell 206, labeled “cell #1.” Despite the use of power control within each cell, out-of-cell terminal units cause interference which is not under the control of the receiving base station within the cell. Thus, for example, power control within cell 203 can be affected by interference from the transmission to terminal units in cell 206 and vice versa.
For example, in a mixed voice and data communication system, transmission of data within cell 203 can cause interference in a neighboring cell such as cell 206. The interference in cell 206 causes increased power allocation to terminal units in cell 206, which is in turn seen as increased interference in cell 203. The increased interference in cell 203 can cause increased power allocation in cell 203, which originally transmitted the data burst. Thus, there is a complete cycle of interaction between the power allocation in cell 203 and cell 206, which resembles a positive feedback loop. The cycle of interaction between the power allocation in cell 203 and cell 206 can lead to higher power consumption than necessary in both cells. The increased power consumption in cell 203 and cell 206 can be seen as increased interference by other neighboring cells, so that the positive feedback effect spreads power control problems from cell 203 and cell 206 to other cells in the system.
An example of feedback effect between two cells only, cell 203 and cell 206, is shown in detail in FIG. 2B and FIG. 2C. FIG. 2B shows graph 230, having power axis 231 plotted against time axis 232. The total transmit power for voice users within cell 203 is shown as aggregate voice power Pv 236 in graph 230. Aggregate voice power Pv 236 varies in time as shown in graph 230. Power for data burst transmission 237 is allocated in addition to aggregate voice power Pv 236. Maximum power limit Pmax 234 that is allocated for the total of aggregate voice and data transmissions in cell 203 is indicated in graph 230 by horizontal solid line Pmax 234.
FIG. 2C shows graph 260, having power axis 261 plotted against time axis 262. Time axis 262 of graph 260 is aligned vertically with time axis 232 of graph 230 so that points on time axis 262 in graph 260 align vertically below the simultaneous points on time axis 232 in graph 230. The total transmit power for voice users within cell 206 is shown as aggregate voice power Pv 266 in graph 260. Aggregate voice power Pv 266 varies in time as shown in graph 260. Maximum power limit Pmax 264 that is allocated for the total of aggregate voice and data transmissions in cell 206 is indicated in graph 260 by horizontal solid line Pmax 264.
Continuing with FIG. 2B and FIG. 2C, graph 230 of FIG. 2B shows that the total transmit power within cell 203 is represented by aggregate voice power curve Pv 236, up until transmission of data burst 237. During data burst 237, the total transmit power within cell 203 is substantially equal to Pmax 234. After data burst 237, the total transmit power within cell 203 is again represented by aggregate voice power curve Pv 236. Similarly, graph 260 of FIG. 2C shows that the total transmit power within cell 206 is represented by aggregate voice power curve Pv 266. As discussed above, the power increase in cell 203, from Pv 236 to approximately Pmax 234, during data burst 237 is seen as increased interference by the users within cell 206. The increased interference in cell 206 leads to higher power allocation by the power control system in cell 206. The higher power allocation is reflected in increase 267 in aggregate voice power curve Pv 266. Conversely, increase 267 in aggregate voice power in cell 206 is seen as increased interference by the users within cell 203 and leads to higher power allocation by the power control system in cell 203. The higher power allocation by the power control system in cell 203 is reflected in increase 238 in aggregate voice power curve Pv 236.
The feedback process continues back and forth between cell 203 and 206 and can lead to a cell allocating the maximum transmit power available, as shown, for example, by maximum 268 in aggregate voice power curve Pv 266. When all available transmit power has been allocated, such as at maximum 268 shown in graph 230 of FIG. 2C, additional users can be denied access to the communication system. To the extent that additional users would have been able to access the communication system, system performance has been degraded. Further the communication link quality for the current users may also be degraded. As pointed out above, the effect can spread from cell to cell and is not restricted to the first pair of cells. Thus, FIG. 2A, FIG. 2B, and FIG. 2C illustrate an example of some of the effects between cells of data transmission on power control in a CDMA or spread spectrum communication system.
As noted above, mixed transmission of voice and data in a CDMA or spread spectrum communication system can subject the system to large swings or variations in the amount of transmission power consumed. For example, such large variation is shown in FIG. 1 by aggregate power curve Pv+d 114. As shown in FIG. 1, Pv+d 114 varies from approximately one half of limit of maximum power Pmax 112 to substantially all of Pmax 112. Such large variation, comprising 50% of the maximum power, would be typical for mixed voice and data communication systems where half of the available power is allocated for voice transmission and half of the available power is allocated for data transmission. As seen in FIG. 1 and in FIG. 2A, FIG. 2B, and FIG. 2C, the large variation can lead to over-allocation and under-allocation of power to each user and to the aggregate of all users within one cell or several cells in the communication system. The resulting instability of power control in the communication system can cause serious degradation of system performance including access problems and degradation of communication link quality for the users.
Thus, there is a need in the art for transmitting mixed voice and data signals without causing abrupt large variations in power consumption. There is also a need in the art for transmitting mixed voice and data signals without causing sudden large reactions in power control. Further, there is need in the art for transmitting mixed voice and data signals without causing undue interference within a cell. Moreover, there is a need in the art for transmitting mixed voice and data signals without causing undue interference between cells.